J A E Stephenson & A D M Walker School of Physics University of KwaZulu-Natal

1. Hanover, May/June 2011 J A E Stephenson & A D M Walker School of Physics University of KwaZulu-Natal (stephens@ukzn.ac.za) Analysis of waves near the magnetopause during a period of FLR activity recorded by the Sanae radar

2. Hanover, May/June 2011 Setting the scene (Part 1) • We are continuing our study of the excitation of Pc5 oscillations in the solar wind driving field line resonances observed by SuperDARN radars • Here we extend our study to the magnetosheath • We present simultaneous observations of data from Cluster 4 and the Sanae radar of oscillations at 2.1 mHz. • Our objective – not yet achieved – is ultimately to follow the propagation of such MHD waves, from the solar wind, through the magnetosheath, to the resonant field line so as to understand the mechanism of energy transfer in detail.

3. Hanover, May/June 2011 Setting the scene (Part 2) • The nature of the driving mechanism of FLRs is an important question in pulsation physics. Different mechanisms may operate at different times. • One mechanism is the Kelvin-Helmholtz instabilty on the magnetopause, excited by the solar wind. The can penetrate the magnetopause and travel as an evanescent fast wave in the magnetosphere. This wave then in turn excites a FLR. This does not explain discrete frequencies. • Cavity modes explain discrete frequencies but long-lived pulsations (many hours) require the cavity to be stable . • Previously (SD 2009,2010) we have presented evidence that FLRs can be driven by a coherent MHD wave in the solar wind. The wave of the appropriate frequency can leak into the waveguide (better analogy than cavity) and then excite a FLR.

4. Hanover, May/June 2011 Earlier result: 07 June 2000 • Top panel: MHD energy flux in solar wind together with amplitude of analytic signals of two velocity components • Middle panel: Energy flux into ionosphere and amplitude of Doppler analytic signal • Bottom panel: Phase differences • Figure published in • Ann. Geophys., 28, 47-59, 2010 • together with figure demonstrating amplitude coherence of better than 97%

5. Hanover, May/June 2011 Doppler velocity in Sanae Beam 4 • Event of 03 June 2006 (10:00-20:00 UT) • Pulsations evident as alternating positive and negative bands in Doppler velocity • Beam 4 (of 16) selected, most closely aligned with lines of magnetic latitude

6. Hanover, May/June 2011 03 June 2006 • Period of maximum 2.1 mHz pulsation activity • Range gate 10 (65.40S AACGM) selected • 10 hour event

7. Hanover, May/June 2011

8. CLUSTER 4 raw data (GSE)Start time 10:00 UT • indicate 2.1 mHz resonance present Step in data accounted for when calculating background field Hanover, May/June 2011

9. Analysis Procedures Hanover, May/June 2011

10. Multi-Taper (Window) Method This method is used to reduce bias due to leakage and to recover lost information that would occur with a single taper. Number of tapers with potentially good bias properties determined by k = 2NWt – 1 Reasonable choice of W=0.08 mHz must take into account trade-off between leakage and variance. Allows for determination of confidence levels against a null hypothesis of a noisy spectrum In addition, the variance of the spectrum can be calculated by jack-knifing, which is achieved by deleting each window in turn from the analysis Hanover, May/June 2011

11. MTM Spectra of Sanae Radar (Beam 4) and Cluster 4 Vy component • 5 Tapers used • W (half width) =0.084mHz • 2.1 mHz peak above 95% significance in radar spectrum • 2.1 mHz peak above 99% significance in Cluster Vy • Common narrowband peaks near 2.1 mHz are shaded. Shading indicates width of peaks (2W) used for complex demodulation. Hanover, May/June 2011

12. 2.1 mHz MTM reconstructed signal Instantaneous amplitude and phase of narrowband resonances determined by method of complex demodulation whereby data were bandpass filtered (in this analysis with the bandwidth of MTM) and an analytic signal was determined Hanover, May/June 2011

13. Coherence between CLUSTER and Sanae radar The diagram shows the coherence of Cluster vy and Sanae Doppler velocity. In the 2.1mHz band it is significant at the 97% confidence level. The work also showed that there was phase coherence between the signals. Hanover, May/June 2011

14. Some Properties of MHD Waves • For 1 four waves exist – fast and slow magnetosonic waves, transverse Alfvén and an entropy wave. • The magnetosonic waves have important contributions from the plasma and magnetic field pressure: the transverse Alfvén wave is incompressible • The fast wave is not highly anisotropic – it is propagated in all directions. Energy in the slow wave is propagated approximately along the magnetic field. Alfvén energy is propagated exactly along the magnetic field for all wave normal directions. • In a stationary medium the wave energy density is • And the wave flux vector is • In the solar wind the wave flux is • V is large enough so that the second term dominates

15. General conditions in the magnetosheath Hanover, May/June 2011

16. Contributions to 2.1mHz Wave energy density Contributions from perpendicular magnetic and kinetic are ANTI-correlated Hanover, May/June 2011

17. Energy flux in CLUSTER rest frame Rest flux dominated by perpendicular components Hanover, May/June 2011

18. Hanover, May/June 2011 Conclusions and future work While we are NOT making the case that this is the only mechanism as the source of FLRs. On previous occasions, we have found discrete oscillations in the Pc5 band that exist in the solar wind are strongly correlated (both in phase and amplitude) with those observed in the magnetosphere. In this case study, they are also found in the magnetosheath. We are performing an in-depth study of this wave in the magnetosheath in order to determine the nature of the wave. Data from CLUSTER 1,2 and 3 spacecraft will be employed to determine further characteristics e.g. wavenumber of the resonance

19. Hanover, May/June 2011 Acknowledgements: We thank members of the SSA-MTM team at the Department of Atmospheric Sciences, UCLA, US Geological Survey and Commissariat a l’Energie Atomique, as well as all other individuals responsible for the development and maintenance of the Toolkit used in the multitaper analysis presented here. We thank members of the Cluster FGM team for supplying the Cluster data. The SHARE radar is supported by the National Research Foundation of South Africa and Antarctic logistics are provided by the Department of Environment Affairs.